A CMOS load problem between Crystal Oscillator and AD630

[yyan71]

Hi there,

I have a problem of realizing signal AM modulation by using a 32kHz Crystal Oscillator (IC) and AD630.

My target is achieving AM modulation by using AD630. In my experiment, the input signal to be modulated is a 2V(Vpp) 20Hz SIN signal, and the carrier input is a 5V(Vpp) 32.768kHz square wave.

Before I build the PCB circuit, I have tried to realize this modulation on breadboard with signal generator as the carrier signal input which is connected to the 50 ohms output on signal generator. The circuit on breadboard works.

Instead of using signal generator, I built a PCB with surface mount Crystal Oscillator as the 32.768kHz carrier input. However, the circuit could not work.

The code of Crystal Oscillator is SG-3030LC. I have checked the datasheet of SG-3030, its output load condition is in CMOS with 15 pF Max.

Now, my questions are:

1. whether AD630's carrier input (PIN 9 of AD630) does not match SG-3030's CMOS load requirement?

2. Is AD630 in a type of TTL or CMOS or what else?

3. Is there a need to put a buffer or CMOS to TTL converter like CD4049 to match them or any other methods suggested?

AD630's datasheet is here:
https://www.analog.com/static/imported-files/data_sheets/AD630.pdf

SG-3030 datasheet is here:
https://www.farnell.com/datasheets/57483.pdf

Please anyone could help. I have got stuck in this problem for a few days and still could not find a solution.

Anyone could be my teacher. I would say thank you here!

Regards

Yang

 

[karesz]

Hi Yang,
Do you have on both supply pins(Nr 1 & 12) the 5V please?
Does it oscillate if the output is separated_loadless?
What are your supply voltages?
The Inputs are Op-Amp like, in all cases high impedance parts...
BUT refer pls. for AD630`s data sheet:
Input ranges (Signal & Comparator!)_they has a smaller range as +/- Vs!! :-( Possible is these your problem...
K.

 

[yyan71]

Re: A CMOS load problem between Crystal Oscillator and AD63

Hi Yang,
Do you have on both supply pins(Nr 1 & 12) the 5V please?
Does it oscillate if the output is separated_loadless?
What are your supply voltages?
The Inputs are Op-Amp like, in all cases high impedance parts...
BUT refer pls. for AD630`s data sheet:
Input ranges (Signal & Comparator!)_they has a smaller range as +/- Vs!! :-( Possible is these your problem...
K.

Hi K,

I have both pins 1 and 12 connected to 5V for SG-3030(Crystal Oscillator).

It can oscillate normally without load. Even if connect the oscillator's output to AD630's pin 9 (carrier input), the oscillated square wave can still be detected on CRO.

Any other suggestions? Thanks K.

Yang

 

[IanP]

Re: A CMOS load problem between Crystal Oscillator and AD63

The SG-3030 is an extremely low-power oscillator with quite low load capacitance ..
There are always PCB considerations while using this stuff ..
I think you should use a buffer but I’d rather go for the CD4050 which suppose to have lower input capacitance than the CD4049 .. see attached picture ..

Rgds,
IanP

 

[FvM]

Re: A CMOS load problem between Crystal Oscillator and AD63

Even if connect the oscillator's output to AD630's pin 9 (carrier input), the oscillated square wave can still be detected on CRO.

What do you mean? Either pin 9 shows a regular oscillator square wave, than it's a problem of AD630 operation or it doesn't, then it's
an oscillator or loading problem.

I guess, the oscillator is operating correct, but you missed to connect pin 10 to a suitable reference voltage. Please consider, why
it can't be correct to connect it to GND.

The SG-3030 is an extremely low-power oscillator with quite low load capacitance ..

It has still a buffered logic output and 15 pF isn't actually an extreme low load impedance specification. I don't expect problems from this side.

 

[yyan71]

Re: A CMOS load problem between Crystal Oscillator and AD63

Even if connect the oscillator's output to AD630's pin 9 (carrier input), the oscillated square wave can still be detected on CRO.

What do you mean? Either pin 9 shows a regular oscillator square wave, that it's a problem of AD630 wiring or it doesn't, then it's
an oscillator or loading problem.

I guess, the oscillator is operating correct, but you missed to connect pin 10 to a suitable reference voltage. Please consider, why
it can't be correct to connect it to GND.

Hi FvM,

I mean oscillator is working properly with an output square wave of 32.768kHz (Vpp=5V).

It is not a wiring problem. I guess it is a CMOS loading problem. Or it is a reference problem.

Please see the attached schematics, the voltages have been labeled for some points. All labeled voltages are referred to true ground which is 0V.

 

[yyan71]

By the way, can I just try to use general OP amplifier as the buffer? Like OP97.

 

[karesz]

Hallo,
You can buffern w. an Op, but it is, in my opinion, more system identical (needs i.e. an SOT23 package & a 10nF, 0603-or smaller) to do it with an Tyni logic LVHC04 or 14 to do...
https://www.fairchildsemi.com/pf/NC/NC7SZ04.html
K.

Added after 13 minutes:

Hej,
You MUSST have a reference voltage on AD630/pin10!
Then, if pin 9 has a right 5V oscill.-signal, what do you wish from your Xtaloosci pls? Its OK!
You can apply a pull up to pin 7 & +Vcc ->> then its a control pin for functionality of your comparator!
Good luck!
K.

 

[FvM]

Re: A CMOS load problem between Crystal Oscillator and AD63

Forget about the CMOS loading problem.

As I mentioned in my previous post, the comparator reference input pin 10 should not be connected to ground respectively
the oscillator negative supply, but a voltage above this value. Pin 9 must swing above and below pin 10 to operate
the comparator input. You should also check the specified voltage range for pin 9. I fear, the oscillator high level is out of
specified voltage range.

 

[karesz]

Hi,
Exactly, & these was my last sentence in the first answer in that topic too ...
K.

 

[FvM]

Re: A CMOS load problem between Crystal Oscillator and AD63

these was my last sentence in the first answer

Yes, I know. I checked the exact specification and found, that the voltage at pin 9 is allowed to swing to the positive supply (now
it's 1 V below), if pin 10 is within the comparator's common mode range. But in any case, it must be above the clock signal's low
level, as said. Biasing pin 10 to 8.5 V by a voltage divider would be the ideal level.

 

[yyan71]

Re: A CMOS load problem between Crystal Oscillator and AD63

Added after 13 minutes:


You MUSST have a reference voltage on AD630/pin10!

Thanks K for your good suggestion.

Maybe the previous attached diagram is not clear. I have attached a new one which clearly labeled the Pin 10 connection. Actually, pin 10/AD630 is connected to the output of TLE2426 voltage reference ground which is 6V in this circuit.

 

[yyan71]

Re: A CMOS load problem between Crystal Oscillator and AD63

Thank you very much Fvm and K, I will try to follow your suggestions and hope it can be solved. Will let you know soon.

 

[karesz]

No-no Yan,
These was for us both with Frank clear.
He wrote explicit and in clear words too, what are the needed voltages on pins 9/10 for sufficient workings possibility of your comparator, he proposed too, you should over think your GND system...
Why is not to use a "normal double supplying" with a GND from that?
In so a system you will have only problems, tan is the input supply to isolate from all other parts in your system, to avoid shorted supply voltages, thinking is complexer, you will everytimes forget, ah these is yet my gnd-not the + 6V and so on.. Your output has the same probmlem, GNdD isNOT GND, for the next instrument!!
Its on + 6V ->>>you must have AC coupling for dont "import" problems over GND-shorts...
Anyway, the op has the needed voltage ranges & differences so & so on Ins, independently from that, if its to GND referenced or from +6V.
Pls read repeatedly the data sheet; Ins have a range not from +/- supply!!
Frank told over that extra too, I only in referencing to datasheet values_you must have on Input signal some V over the minus supply & some V under the + supply voltage!
Sorry I can not read yet the data sheet & I dont remember for exact values, but read it pls , its at 1.5--2V less from both direction of supply voltage values, also an input pp range of ca. 3V maximum!
K.

 

[yyan71]

Thanks Karesz and Fvm,

Your suggestions help a lot!

After adding a buffer between Crystal Oscillator and AD630 with the buffer's positive input pulled down to a proper reference, the problem has been solved. I learned a lot.

Thank you!! ^_^

Yang